SEARCH

SEARCH BY CITATION

Keywords:

  • Death;
  • Epilepsy;
  • Lamotrigine

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Supporting Information

Purpose:  Nonrandomized studies of the relationship of antiepileptic drugs (AEDs) with sudden unexpected death in epilepsy (SUDEP) may be susceptible to confounding by tonic–clonic seizure frequency, polypharmacy, and other potential risk factors for SUDEP. We evaluated the risk of SUDEP with lamotrigine (LTG) compared to active comparators and placebo in randomized controlled clinical trials conducted by GlaxoSmithKline (GSK) between 1984 and 2009.

Methods:  Among 7,774 subjects in 42 randomized clinical trials, there were 39 all-cause deaths. Ten deaths occurred >2 weeks after discontinuation of study medication and were excluded. Narrative summaries of deaths were independently reviewed by three clinical experts (TT, LH, DF), who were blinded to randomized treatment arm. The risk of definite or probable SUDEP was compared between treatment arms for each trial type (placebo-controlled, active-comparator, crossover), using exact statistical methods.

Key Findings:  Of 29 on-treatment deaths, eight were definite/probable SUDEP, four were possible SUDEP, and 17 were non-SUDEP. The overall, unadjusted rate of definite/probable SUDEP for LTG was 2.2 events per 1,000-patient years (95% confidence interval [95% CI] 0.70–5.4). The odds ratios (OR) for on-treatment, definite/probable SUDEP in LTG arms relative to comparator arms, adjusted for length of exposure and trial, were the following: placebo-controlled, OR 0.22 (95% CI 0.00–3.14; p = 0.26); active-comparator, OR 2.18 (95% CI 0.17–117; p = 0.89); and placebo-controlled cross-over, OR 1.08 (95% CI 0.00–42.2; p = 1.0).

Significance:  There was no statistically significant difference in rate of SUDEP between LTG and comparator groups. However, the CIs were wide and a clinically important effect cannot be excluded.

Sudden unexpected death in epilepsy (SUDEP) can occur in generally healthy people with epilepsy, often in the context of a generalized tonic–clonic seizure (Shorvon & Tomson, 2011). In a meta-analysis of SUDEP, the mean incidence of SUDEP in patients with epilepsy was 2.17 per 1,000 person-years, with a wide range (0.1–9.0 per 1,000 person-years) depending on age and epilepsy severity; SUDEP is responsible for approximately 17% of epilepsy deaths (range 4–40%) (Hughes, 2009). The most important risk factor for SUDEP is reported to be uncontrolled generalized tonic–clonic seizures. Subtherapeutic anticonvulsant drug levels, age (young adults), antiepileptic drug (AED) polytherapy, lack of AED treatment, male sex, younger age at onset, and longer duration of epilepsy have also been reported as possible risk factors (Nilsson et al., 1999; Tomson et al., 2008; Hughes, 2009; Hesdorffer et al., 2011).

In most studies, SUDEP has not been associated with the use of any particular AED (Tomson et al., 2005, 2008), although carbamazepine (CBZ) (Timmings, 1993; Langan et al., 2005) and lamotrigine (LTG) (Hesdorffer et al., 2011; Aurlien et al., 2007) have been reported to be associated with a higher risk of SUDEP in some nonrandomized studies. A Norwegian case series of SUDEP was supplemented by a case–control comparison suggesting that LTG was associated with increased incidence of SUDEP in female patients (Aurlien et al., 2012). Such nonrandomized studies of the relationship of AEDs with SUDEP may, however, be susceptible to confounding by tonic–clonic seizure frequency, polypharmacy, epilepsy syndrome, and other potential risk factors for SUDEP.

In a pooled analysis of four case–control studies of SUDEP, LTG was associated with an odds ratio (OR) of 1.86 (95% confidence interval [95% CI] 1.22–1.84) for SUDEP relative to no AED treatment, adjusted for age, sex, data source, and duration of epilepsy (Hesdorffer et al., 2011). However, with further adjustment for generalized tonic–clonic seizure frequency, LTG was not associated with a significantly greater risk of SUDEP: The OR for SUDEP relative to no AED treatment was 0.7 (95% CI 0.1–3.6) for LTG monotherapy, and 0.95 (95% CI 0.4–2.2) for LTG polytherapy (Hesdorffer et al., 2012).

Randomized clinical trials provide the most scientifically rigorous approach for evaluating the effect of drug therapies. Among LTG-exposed patients in double-blind as well as open-label nonrandomized studies through 1992 (N = 4,700 patients, 5,747 person-years of exposure), 18 definite or highly probable cases of SUDEP were identified by adjudication. The rate of definite/highly probable SUDEP was 2.78 per 1,000 patient-years of exposure to LTG (Leestma et al., 1997). The current analysis of data from all available GlaxoSmithKline (GSK)–controlled clinical trials, which includes some patients from the earlier analysis (but not those from open-label or named patient studies), was conducted primarily to compare the incidence of SUDEP for epilepsy patients treated with LTG, placebo, or other active comparators using specified criteria to define SUDEP (Annegers, 1997), as classified by a formal adjudication panel.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Supporting Information

Study population

The study population consisted of all 7,774 subjects in 42 GSK-sponsored randomized, controlled clinical trials of LTG for epilepsy from 1984 to 2009 (Table S1). Data were unavailable from 2 of 44 GSK-controlled studies, despite efforts to locate the original data. Included trials were 13 parallel-group placebo-controlled trials (N = 1,916 subjects/exposures), 17 parallel-group active-controlled trials (N = 5,369 subjects/exposures), and 12 crossover placebo-controlled trials (N = 489 subjects; 933 exposures). Two trials with pseudo-placebo comparator groups are included in the placebo group analysis. The studies included adult and pediatric patients with newly diagnosed epilepsy, refractory epilepsy, partial seizures, and generalized seizures. All clinical trial sites obtained approval from their local institutional review boards. Written informed consent was obtained from study participants, legally authorized representatives, or both, according to local guidelines.

Data collection

The following subject-level data were extracted retrospectively, where available, from each study: study number, study type (placebo-controlled parallel-group, placebo-controlled crossover, active-controlled parallel group), study treatment group, age, sex, race, baseline frequency of seizures, baseline seizure types, concurrent AEDs, number of AEDs, duration of epilepsy, study treatment start date, study treatment stop date, death (yes/no), and date of death. Low-dose active and placebo exposures were classified as placebo, since the low doses were considered to be nonefficacious doses. Low-dose active exposure was chosen as a pseudo-placebo intended to protect subjects from severe exacerbations of seizure activity, although not providing optimal treatment.

SUDEP classification

All deaths in the included GSK-sponsored clinical trials for LTG were identified using the GSK global clinical safety database and original study reports. Case narratives for subjects who died were extracted from the GSK global clinical safety database and from study reports; these were supplemented with additional information available from original study case-report forms and autopsy reports, and the treatment arm (trial drug) information was blinded.

SUDEP classification was adjudicated by a panel of experts (DF, LH, TT), who agreed on the objectives, common terminology, and definition of SUDEP to be applied to the adjudication process. They independently reviewed all treatment-blinded case narratives and autopsy reports, where available, and classified the deaths using both Annegers (1997) and Nashef et al. (2012) criteria. They then agreed on a final classification at a consensus meeting.

Statistical analysis

Because the relative risk of SUDEP on LTG versus comparator arms may differ based on whether the comparator is placebo or another AED, or if there are carryover effects in crossover studies, analyses were performed separately for each study type.

The primary outcome measure was the incidence of on-treatment definite and probable SUDEP according to Annegers criteria (Annegers, 1997). On-treatment SUDEP was defined as SUDEP that occurred ≤2 weeks after discontinuation of trial medication (by which time LTG or the active comparator would have been cleared from the circulation). A secondary outcome measure was definite/probable/possible SUDEP. All-cause death was examined in a post hoc analysis. Person-time was calculated from time of randomization until the earlier of (1) end of study; (2) withdrawal from study; or (3) death. Exact 95% confidence intervals (CIs) were calculated for the incidence rates.

Exact logistic regression was used to estimate the OR and 95% CIs of LTG versus comparator adjusted for trial and length of exposure.

Post hoc analyses estimated the incidence of SUDEP and calculated the ORs for LTG compared to (1) pure placebo arms excluding low-dose active pseudo-placebo treatments, and (2) active comparators excluding CBZ, which has previously been reported to be associated with an increased risk of SUDEP (Timmings, 1993; Langan et al., 2005).

A post hoc analysis calculated the unadjusted rate of SUDEP in studies of (1) newly diagnosed patients and (2) patients with refractory epilepsy.

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Supporting Information

The characteristics of the 7,774 subjects stratified by study type are presented in Table 1. In the active-controlled trials, the active comparators were carbamazepine (42%), valproate (40%), levetiracetam (6%), phenytoin (4%), topiramate (4%) and zonisamide (4%). The total treatment exposure per study type was 416 person-years for LTG and 294 person-years for placebo or low-dose active comparator within the parallel-group placebo-controlled studies, 1,285 person-years in the LTG group and 960 person-years in the active group within parallel-group active-controlled studies, and 106 person-years for LTG and 107 person-years for placebo or low-dose active within crossover placebo controlled studies.

Table 1.   Demographic characteristics of the 42 studies by study type
CharacteristicPlacebo-controlled parallel-group studiesActive-controlled parallel-group studiesPlacebo-controlled crossover studiesa
Lamotrigine (N = 1,093)Placebo (N = 823)Lamotrigine (N = 3,061)Active Comparator (N = 2,308)Lamotrigine (N = 477)Placebo (N = 456)
  1. aNote that the placebo-controlled crossover studies have the majority of the same patients in both arms, although some patients only received one treatment.

  2. SD, standard deviation.

Age (years)(N = 1,085)(N = 819)(N = 3,060)(N = 2,308)(N = 452)(N = 432)
 Mean ± SD29.0 ± 13.9726.8 ± 15.0532.1 ± 20.232.7 ± 19.132.1 ± 13.432.2 ± 13.6
 Median (range)29.0 (0–73)26.0 (1–74)27.0 (2–94)29.0 (2–88)31.0 (7–67)31.0 (7–67)
Sex, N (%)(N = 1,085)(N = 819)(N = 3,061)(N = 2,308)(N = 452)(N = 432)
 Female546 (50)411 (50)1,664 (54)1,222 (53)215 (48)202 (47)
 Male539 (50)408 (50)1,397 (46)1,086 (47)237 (52)230 (53)
Race, N (%)(N = 1,084)(N = 819)(N = 1,386)(N = 1,153)(N = 226)(N = 216)
 African-American85 (8)65 (8)57 (4)68 (6)10 (4)11 (5)
 Asian160 (15)152 (19)181 (13)171 (15)4 (2)4 (2)
 Hispanic17 (2)21 (3)35 (3)38 (3)0 (0)0 (0)
 White-Caucasian786 (73)537 (66)1,086 (78)868 (75)207 (92)196 (91)
 Other36 (3)44 (5)27 (2)8 (<1)5 (2)5 (2)
Seizure duration (years)(N = 1,054)(N = 792)(N = 790)(N = 686)(N = 340)(N = 325)
 Mean ± SD18.1 ± 11.616.5 ± 11.97.3 ± 10.38.4 ± 11.519.5 ± 12.219.3 ± 12.2
 Median (range)17.0 (0–59)14.0 (0–68)3.0 (0–56)3.0 (0–60)18.1 (0–62)18.1 (0–62)
Baseline seizure types, N (%)(N = 1,064)(N = 800)(N = 2,546)(N = 2,015)(N = 375)(N = 355)
 Partial829 (78)543 (68)1,285 (50)1,085 (54)352 (94)334 (94)
 Generalized224 (21)245 (31)1,050 (41)730 (36)11 (3)12 (3)
 Both11 (1)12 (2)211 (8)198 (10)12 (3)9 (3)
Baseline seizure frequency per month(N = 651)(N = 597)(N = 2,054)(N = 1,514)(N = 366)(N = 344)
 Mean ± SD34.7 ± 122.334.3 ± 111.816.8 ± 83.322.6 ± 260.129.6 ± 107.830.3 ± 111.0
 Median (range)9.4 (1–2,258)8.9 (1–2,165)1.8 (0–1,721)2.0 (0–9,604)10.6 (1–1,840)11.0 (1–1,840)
Any concomitant AED medication(s), N (%)(N = 1,073)(N = 813)(N = 1,083)(N = 857)(N = 400)(N = 380)
 Acetazolamide16 (1)10 (1)0 (0)1 (<1)1 (<1)1 (<1)
 Barbiturate217 (20)122 (15)35 (3)34 (4)80 (20)78 (21)
 Benzodiazepine193 (18)160 (20)60 (6)47 (5)41 (10)41 (11)
 Carbamazepine611 (57)402 (49)430 (40)404 (47)295 (74)279 (73)
 Ethosuximide17 (2)16 (2)10 (<1)4 (<1)3 (<1)3 (<1)
 Felbamate1 (<1)0 (0)1 (<1)0 (0)0 (0)0 (0)
 Gabapentin6 (<1)7 (<1)18 (2)7 (<1)0 (0)0 (0)
 Levetiracetam22 (2)17 (2)1 (<1)0 (0)0 (0)0 (0)
 Oxcarbazepine20 (2)27 (3)32 (3)26 (3)9 (2)6 (2)
 Phenytoin361 (34)246 (30)300 (28)253 (30)123 (31)116 (31)
 Pregabalin1 (<1)0 (0)0 (0)0 (0)0 (0)0 (0)
 Tiagabine2 (<1)0 (0)0 (0)0 (0)0 (0)0 (0)
 Topiramate36 (3)33 (4)8 (<1)3 (<1)0 (0)0 (0)
 Vigabatrin22 (2)30 (4)6 (<1)6 (<1)19 (5)17 (4)
 Valproic acid243 (23)255 (31)296 (27)186 (22)84 (21)76 (20)
 Zonisamide30 (3)28 (3)5 (<1)1 (<1)0 (0)0 (0)
 Other10 (<1)14 (2)8 (<1)4 (<1)0 (0)0 (0)
 Unclear1 (<1)1 (<1)0 (0)2 (<1)2 (<1)2 (<1)
Days on treatment(N = 1,093)(N = 823)(N = 3,023)(N = 2,285)(N = 442)(N = 451)
 Mean ± SD139.9 ± 50.1131.6 ± 44.3156.1 ± 99.1154.4 ± 107.688.1 ± 24.887.9 ± 24.7
 Median (range)142.0 (2–266)133.0 (1–217)167.0 (1–538)140.0 (1–545)84.0 (7–133)84.0 (7–139)

There were 39 all-cause deaths; 10 deaths occurred >2 weeks after discontinuation of study medication and were excluded from the analysis of on-treatment events (Table S2). Of the 29 on-treatment deaths, 8 were eight definite or probable SUDEP (four in LTG, four in comparators), 4 were possible SUDEP, and 17 were non-SUDEP, as defined by Annegers criteria.

The pooled rate of definite/probable SUDEP for LTG was 2.2 events per 1,000-person years (CI 0.70–5.3). Rates in the pooled LTG-exposed calculations are not adjusted for study type or other factors. The pooled rates by sex, epilepsy type, and seizure type are summarized in Table 2.

Table 2.   Crude, pooled rates for definite or probable SUDEP in patients exposed to lamotrigine, combined for all study types but not adjusted for study type or other factors
Study population subgroupsNumber of patientsSUDEP rate per 1,000 person-years (95% CI)
  1. aSex missing for 61 patients.

  2. bThere were no SUDEP cases in the studies within LTG patient populations designated as “well-controlled” (n = 50), combined newly diagnosed and refractory patients (n = 92), or “not available” (n = 310).

  3. cThere were no SUDEP cases among the patients with only generalized seizures or with information unavailable (n = 647 and 91, respectively).

  4. dGeneralized seizures included: absence, myoclonic, clonic, tonic, tonic–clonic, and atonic.

  5. ePartial seizures included simple partial, complex partial, and secondarily generalized.

All patients exposed to LTG4 SUDEP/4,631 patients2.22 (0.70–5.3)
Sexa  
 Male2 SUDEP/2,173 patients2.5 (0.42–8.2)
 Female2 SUDEP/2,425 patients2.0 (0.34–6.6)
Epilepsy populationb  
 Newly diagnosed2 SUDEP/1,695 patients2.5 (0.42–8.2)
 Refractory2 SUDEP/2,484 patients2.2 (0.38–7.4)
Seizure typec  
 Generalizede0 SUDEP/647 patients N/A
 Partialf2 SUDEP/2,032 patients2.7 (0.45–8.8)
 Both generalized and partial2 SUDEP/1,861 patients2.4 (0.40–8.0)

The ORs for on-treatment, definite or probable SUDEP events in the LTG arms relative to the comparator arms, adjusted for length of exposure and trial, were not significant: parallel-group placebo-controlled trials OR 0.22 (95% CI 0.00–3.14; p = 0.26); parallel-group active-controlled trials OR 2.18 (95% CI 0.17–117; p = 0.89); placebo-controlled crossover trials OR 1.08 (95% CI 0.00–42.2; p-value = 1.0) (Table 3). In a subanalysis of the parallel-group, placebo-controlled trials excluding the low-dose active pseudo-placebo arms, the OR for SUDEP comparing LTG to pure placebo was 0.34 (0.00–13.1). In a subanalysis of the parallel-group active-controlled trials excluding carbamazepine arms, the OR for SUDEP comparing LTG to non-CBZ active comparators was 1.80 (0.13–∞).

Table 3.   SUDEP frequency, rates, and odds ratios comparing lamotrigine treatment group and comparator, by study type
Study typeStudies (n)LTG n/N Rate (95% CI)Comparator n/N Rate (95% CI)OR (95% CI) p-Valuea
  1. aThe odds ratio (OR) relative to placebo adjusted for exposure and trial for definite or probable SUDEP, by study type. p-Value tests hypothesis that OR is 1.

Definite or probable SUDEP    
 Parallel-arm, placebo-controlled   Rate per 1,000 person-years (exact 95% CI)130 SUDEP/1,093 patients 0.00 (0.00–8.83)2 SUDEP/823 patients 6.78 (0.82–24.27)0.22 (0.00–3.14) p = 0.26
 Parallel-arm, active comparator   Rate per 1,000 person-years (exact 95% CI)174 SUDEP/3,061 patients 3.11 (0.85–7.95)1 SUDEP/2,308 patients 1.04 (0.03–5.78)2.18 (0.17–117) p = 0.89
 Crossover studies   Rate per 1,000 person-years (exact 95% CI)120 SUDEP/477 patients 0.00 (0.00–34.2)1 SUDEP/456 patients 9.26 (0.23–50.51)1.08 (0.00–42.2) p = 1.00
Definite, probable or possible SUDEP    
 Parallel-arm, placebo-controlled   Rate per 1,000 person-years (exact 95% CI)130 SUDEP/1,093 patients 0.00 (0.00–8.83)2 SUDEP/823 patients 6.78 (0.82–24.27)0.22 (0.00–3.14) p = 0.26
 Parallel-arm, active comparator   Rate per 1,000 person-years (exact 95% CI)177 SUDEP/3,061 patients 5.45 (2.19–11.19)2 SUDEP/2,308 patients 2.08 (0.25–7.50)2.03 (0.36–20.8) p = 0.63
 Crossover studies   Rate per 1,000 person-years (exact 95% CI)120 SUDEP/477 patients 0.00 (0.00–34.2)1 SUDEP/456 patients 9.26 (0.23–50.51)1.08 (0.00–42.2) p = 1.00

Results were similar for the secondary outcome of definite/probable/possible SUDEP. There were no possible SUDEP cases in the parallel-group placebo-controlled trials, or in the placebo-controlled cross-over trials. For the parallel-group active-controlled trials, there were seven definite/probable/possible SUDEP cases in 3,061 LTG-exposed patients, corresponding to a rate of 5.45 per 1,000 person years (95% CI 2.19–11.19); there were two definite/probable/possible cases among 2,308 patients in the active comparator arms, corresponding to a rate of 2.08 per 1,000 person-years (95% CI 0.25–7.50). The OR comparing definite/probable/possible SUDEP rates in LTG and the comparator arm was 2.03 (95% CI 0.36–20.8, p = 0.63).

Sex-stratified ORs of definite/probable SUDEP for LTG versus comparator in each study type were limited by the low event rates. The OR for LTG versus placebo among male participants in placebo-controlled trials was 0.20 (95% CI 0.00–2.73, p = 0.225). For female participants this could not be computed due to zero events in both treatment groups. The OR for LTG versus active comparator among male subjects in active-controlled trials was 0.65 (95% CI 0.01–53.9, p = 1.000); the OR for female subjects was 1.96 (95% CI 0.15–∞, p = 0.611). The OR for LTG versus placebo among female subjects in crossover trials was 1.15 (95% CI 0.0–45.0, p = 1.000). For male subjects this could not be computed due to zero events in both treatment groups.

There was no significant association of LTG with all-cause death, with ORs of LTG versus comparators of 0.17 (95% CI 0.00–1.68, p = 0.13) in parallel-group placebo-controlled trials; 0.93 (95% CI 0.37–2.50, p = 1.00) in active-controlled trials; and 1.08 (95% CI 0.00–42.2, p = 1.00) in placebo-controlled crossover trials.

We also compared the concordance of Annegers criteria with the new SUDEP classification by Nashef et al. (Table 4). One probable SUDEP under the Annegers criteria was reclassified as a definite-plus case because of a potential competing cause of death. Of six definite SUDEP classifications, two changed to definite-plus with the new criteria. Of four possible SUDEP classifications, one changed to unclassified. All non-SUDEP classifications were in agreement between the two sets of criteria.

Table 4.   Concordance of Annegers (1997) and Nashef et al. (2012) criteria for SUDEP
 Annegers classification
Definite (n = 6)Probable (n = 2)Possible (n = 4)Not SUDEP (n = 17)
  1. aThe adjudicators discussed this case in detail and decided that it was either a “possible” or “probable” case with the Annegers classification. An autopsy was performed in this case; however, the patient did not die “in a reasonable state of health,” and there was a competing cause present. Therefore, the case was classified as probable. The new Nashef et al. criteria require only a reasonable state of health in the probable category, but not in definite. For the new criteria, a contributing cause (cardiomyopathy) was present so the classification becomes Definite Plus.

Nashef et al. classification    
 Definite (n = 4)4   
 Definite-plus (n = 3)2 1a  
 Probable (n = 1) 1  
 Possible (n = 3)  3 
 Unclassified (n = 1)  1 
 Not SUDEP (n = 17)   17

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Supporting Information

In the clinical development program for LTG, the total number of definite/probable SUDEPs in 42 controlled clinical trials for epilepsy was low (N = 8). The rate of SUDEP in the LTG-exposed arms was 2.2 events per 1,000-person years (CI 0.7- 5.4), consistent with that reported by Leestma in the overall LTG program for epilepsy (including double-blind, open-label, and compassionate use studies; Leestma, 1997) and other AED development programs (Walczak, 2003). The rate of SUDEP in female patients in the LTG exposed arms was similar to that in male patients. The point estimates of the OR for definite/probable SUDEP in LTG versus comparator arms ranged from 0.22 to 2.18, depending on the study type and comparator, with wide confidence intervals including the null. Subanalyses comparing LTG to pure placebo, excluding low-dose active comparators, and to active comparators besides CBZ did not significantly change the OR estimates for definite/probable SUDEP. CBZ has previously been reported to be associated with an increased risk of SUDEP. There was no association of LTG with all-cause death.

The study should be interpreted in the context of its strengths and limitations. The main strengths of the analysis were the relatively large number of patients with LTG exposure and comparator populations, the randomized allocation of patients in the controlled studies (and self-controlled design for the crossover studies), the systematic collection of adverse event data in clinical trials, and the independent expert clinical adjudication for SUDEP of all deaths in the studies.

However, the trial populations are not necessarily representative of the general epilepsy population, follow-up time is limited, and the event numbers for a rare outcome like SUDEP were low. The effect of other risk factors could not be evaluated due to the small number of SUDEP cases in the study. However, the distribution of risk factors at baseline in the clinical studies would not be expected to differ between treatment arms due to the randomized or self-controlled (for crossover) design of the studies.

This study is the first to apply the recently proposed SUDEP diagnostic criteria by Nashef, and compare to the Annegers classification. Although numbers were small, there was concordance between the definite, probable, possible, and not-SUDEP classifications. The incorporation of a “plus” category captures preexisting conditions that could have contributed to a death that was otherwise classified as SUDEP (Nashef et al., 2012). In our study, three of eight definite/probable SUDEP cases according to Annegers criteria were classified as definite-plus according to the Nashef criteria, highlighting the need to better understand the impact of comorbidities on SUDEP risk.

In conclusion, although there was no statistically significant different in rate of SUDEP between LTG and the comparator groups, CIs were wide and a clinically important effect cannot be excluded. A recent meta-analysis of AED clinical trials concluded that adjunctive AEDs at efficacious doses may be associated with a reduced incidence of definite or probable SUDEP compared with placebo (Ryvlin et al., 2011). The point estimate in the current study of the OR for SUDEP in the LTG versus placebo arm (0.22) is in line with the risk reduction reported in the meta-analysis. Large observational studies, including case–control studies of SUDEP, epilepsy registries, or SUDEP registries, can provide additional information on the risks and benefits of specific AEDs on SUDEP. Such studies must carefully capture and control for confounders that might be related both to AED selection and risk of SUDEP, such as the frequency of generalized tonic–clonic seizures, polytherapy, and epilepsy syndrome (Hesdorffer et al., 2012).

Disclosure

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Supporting Information

TT, LH, and DF served as paid consultants for GSK for their role in blinded adjudication of SUDEP events, but they were not paid for authorship activities related to this article. NB, AH, MI, LI, AK, TS, AW, and RL are current employees of GSK. DF receives salary support from The Epilepsy Study Consortium, a nonprofit organization dedicated to improving the lives of patients with epilepsy, and devotes 10% of his time to work done for the Consortium. NYU receives a fixed amount from the Epilepsy Study Consortium towards his salary. The Consortium receives payments from a large number of pharmaceutical companies for consulting activities. Because there are so many companies contributing, the amount from each company contributed toward Dr. Friedman’s salary is minimal. Within the past year, the Epilepsy Study Consortium received payments from the 21 companies listed below. All payments are reported annually and reviewed by NYU’s Conflict of Interest Committee. Cyberonics, Cypress Biosience, Inc., Eisai Medical Research, Entra Pharmaceuticals, GlaxoSmithKline, Icagen, Inc., Intranasal/Ikano, Johnson & Johnson, Marinus, Neurotherapeutics, NeuroVista Corporation, Ono Pharma U.S.A., Inc., Ovation/Lundbeck, Pfizer, Sepracor, SK Life Science, Supernus Pharmaceuticals, Taro, UCB Inc./Schwarz Pharma, Upsher Smith, and Valeant. DF also receives royalties from the sale of What Do I Do Now: Epilepsy (Oxford University Press, 2011) and research support from the American Epilepsy Society. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. Disclosure
  7. References
  8. Supporting Information

Table S1. List of studies.

Table S2. SUDEP patient table.

FilenameFormatSizeDescription
epi3689_sm_TableS1.doc77KSupporting info item
epi3689_sm_TableS2.xlsx15KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.